New Delhi: Social media platforms and other online intermediaries must remove unlawful AI-generated content within three hours after receiving a valid
court order or a reasoned notice from an authorised government agency. The earlier deadline was 36 hours.
The updated rules took effect after the government amended the Information Technology Rules, 2021, on February 10, 2026. Union Minister of State for Electronics and Information Technology Jitin Prasada shared the details in the Lok Sabha on August 5.
AI-generated content must carry clear labels
The amendments cover synthetically generated information, including deepfake audio, videos, images and text. Platforms must place clear labels and traceable metadata on permitted AI-generated content so that users can identify altered material.
Intermediaries must use suitable technical systems, including automated tools, to stop users from creating or sharing AI-generated content that breaks Indian laws.
The rules specifically cover harmful material such as:
- Child sexual exploitation content
- Non-consensual intimate images
- Impersonation through AI
- Misleading and defamatory deepfakes
- Content that threatens public order or national security
Complaint deadlines reduced
Platforms must act faster on user complaints. The deadline for handling complaints involving nudity, impersonation and related material has fallen from 72 hours to 36 hours.
For sensitive cases, the response time has been cut from 24 hours to two hours. Users who are unhappy with a platform’s decision can appeal through the Grievance Appellate Committee portal.
Failure to follow these duties may result in intermediaries losing their legal protection under Section 79 of the Information Technology Act.
Deepfake detection projects receive support
Under the IndiaAI Mission, the government has approved 13 projects linked to responsible AI, including tools for detecting manipulated content.
These projects include Saakshya, developed by IIT Jodhpur and IIT Madras, AI Vishleshak for audio-visual forgery detection and IIT Kharagpur’s real-time voice deepfake detection system.
Citizens can report deepfakes, financial fraud and other cybercrime through the National Cyber Crime Reporting Portal or by calling helpline number 1930.
The government said its aim is to create an “open, safe, trusted and accountable cyberspace for users.” The tighter timelines place greater pressure on platforms to review official notices quickly and improve systems used to detect harmful synthetic media.
